Data Visualization - Tennis

Maxwell Fritz

Data Summary

Weekly Rankings of Top 1000 - Data ranges from 1973-2018
- 2,746,003 obs. of 4 variables
- 1,670,401 obs. of 5 variables
Player Data
- 53,319 obs. of 6 variables
- 21,750 obs. of 6 variables

https://maxfritz.shinyapps.io/rankings/

Weekly Top 10 Graphic

Weekly Top 10 Graphic (cont.)

Top 10 players at given week
Ordered by rank
Height of bars: points
Built with ggplot and gganimate

Weekly Top 10 Graphic (cont.)

Focus on relative positioning
Smooth shifting bars
Auto-scaling x-axis
Label positioning, tracking with variable values

Weekly Top 10 Graphic (cont.)

Small formatting errors
Difficulty in animating transition states between integer values

Weekly Top 10 Graphic (cont.)

Rebuild in D3
- Animation quality
- Proper graphics scaling
- Able to manipulate parameters, start/stop

Ranking History - Step Graph

https://maxfritz.rstudio.cloud/a57643922f924b45bf424d906132dd16/file_show?path=%2Fcloud%2Fproject%2Fatp_no1_hist.html

Ranking History - Step Graph (cont.)

Built with Plotly
Stepped line graph by rank, over time
Ranking history for every World No.1

Ranking History - Step Graph (cont.)

Ability to select and compare individual players
Refine by time frame
Auto-scaling axis
Free of clutter

Ranking History - Step Graph (cont.)

Difficult to differentiate between lines
Coloring scheme difficult to control
Easy to lose focus
Try to implement focus on hover
Edit color scales so that adjacent players have distinct coloring schemes

Choropleth

https://plot.ly/~maxwellfritz/25/

Choropleth (cont.)

Players by country in given year
When adjacent, meant to paint broad picture of player origin over time
Information on hover
Log scaled color so that large countries do not dominate

Choropleth (cont.)

The USSR issue
Add change over time

Next Steps

Small formatting issues
Centralize on Shiny app
Begin working with match-level data

Questions?